Can an attribute ID be arbitrarily picked ?
Have noticed all OOTB attribute ID's come in a long string in the form of
com.ibm.team.workitem.attribute.<attributeName>
like the ID of any other RTC entities.
But when creating new customized attributes, found that their ID can be arbitrarily picked, as long as there are no embedded blanks (and probably most of the special chars).
Is the long, descriptive ID built that way just to have an effective way to actually identify the owner/creator and the type of the entity, i.e. the <attributeName> is a workitem.attribute created by com.ibm.team ?
Is there any implication not to use the implied convention, or to use the convention, besides the creation or the avoidance of a confusion in trying to figure out what an entity is by name ?
Accepted answer
Hi Long,
There is no specific rule regarding that. I have found many repositories where admins just gave id without any prefix. Anyway, from my perspective it could be a good practice to use a convention like for Java packages - give creator name as a prefix to improve search experience when administering attributes.
Let us know if that helps.
There is no specific rule regarding that. I have found many repositories where admins just gave id without any prefix. Anyway, from my perspective it could be a good practice to use a convention like for Java packages - give creator name as a prefix to improve search experience when administering attributes.
Let us know if that helps.
Comments
we override and replace the generated id for custom items with our own
com.company.type
com.xx.workitem.attribute.type.xxx
com.xx.workitem.enumeration.literal.xxx
but thanks for asking.. your question makes me wonder if we have made a mistake on one area. we assigned specific state ids to our custom defect workflow.
but if a workitem changes type, then the states won't map to the new workitem type ids.
you can change a workitem type very easily.
luckily we have only deployed one project on the new process template.